Changing ExecutePipe.cpp code to set scoreboard based on destination
register, not the pipe register.
Added associated test for checking correct register file's rename
registers are used.

Summary of Changes:
LSU.cpp -> Operand dependency checking based on source bits in
scoreboard, callbacks
ExecutePipe.cpp -> Operand dependency checking based on source bits in
scoreboard, callbacks
Rename_test.cpp -> Added testing for both LSU and ExecutePipe with RAW
instruction jsons, imported OlympiaSim for full simulation run to access
data

Final portion of operand dependency checking implemented in this PR for
LSU and ExecutePipe.

---------

Signed-off-by: Aaron Chan <achn357@gmail.com>
Co-authored-by: Knute Lingaard <knute.lingaard@sifive.com>

Signed-off-by: Knute <56927209+klingaard@users.noreply.github.com>

Implementation of Task 1 for full Rename support:

Task 1: Add a arch map -> PRF map to Rename. 

This includes adding a physical register freelist (parameterized to the number). 
When an instruction enters rename, its source arch registers are mapped to the current
physical register for that arch value. Then its destination registers are renamed — 
a new PRF is pulled from the freelist, given to the instruction (via the scoreboard mask) 
and the rename map updated.

---------

Signed-off-by: Aaron Chan <achn357@gmail.com>
Co-authored-by: Aaron Chan <aaronchan@client-10-229-46-41.tamulink.tamu.edu>
Co-authored-by: Aaron Chan <aaronchan@Aarons-MacBook-Air-2.local>
Co-authored-by: Knute Lingaard <knute.lingaard@sifive.com>
Co-authored-by: Aaron Chan <aaronchan@Aarons-Air-2.attlocal.net>
Co-authored-by: Aaron Chan <aaronchan@client-10-228-106-204.tamulink.tamu.edu>

Signed-off-by: Suraj Shirvankar <surajshirvankar@gmail.com>

Signed-off-by: Suraj Shirvankar <surajshirvankar@gmail.com>

This fixes #35 and tracks the changes in Sparta
[PR397(https://github.com/sparcians/map/pull/397

), which are merges into Sparta main already. 

---------

Signed-off-by: Peter Debacker <peter-d@users.noreply.github.com>
Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>
Co-authored-by: Peter Debacker <pdback@imec.be>
Co-authored-by: Knute Lingaard <Knute.Lingaard@sifive.com>
Co-authored-by: Knute Lingaard <klingaard@gmail.com>

Small to change to  use all the CPU cores for the build.

Signed-off-by: Suraj Shirvankar <surajshirvankar@gmail.com>

Added directions to checkout map_v1.1.0

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>

the path of WeightedContextCounter has been changed to
sparta/statistics, so fix it in this PR

First commit of using GitHub actions

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>

---------

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>
Co-authored-by: bdutro <brett.dutro@gmail.com>

Added specification for conda environment which can be used to build
olympia, stf lib/tools

- unmodified patch for dromajo, when applied, did not compile for
      me. Added a missing header file.

Added link to instructions to install required STF library packages

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>

Signed-off-by: Knute <56927209+klingaard@users.noreply.github.com>

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>

Adding a cast to get past the conversion error encountered during build.

Added a few additional LSU statistics. 

Output from running dhrystone trace:

```
        top.cpu.core0.lsu
            lsu_insts_dispatched                              = 5850001
            stores_retired                                    = 5850001
            lsu_insts_issued                                  = 11700073
            lsu_insts_completed                               = 5849999
            lsu_flushes                                       = 0
            tlb_hits                                          = 5850001
            tlb_misses                                        = 16
            dl1_cache_hits                                    = 5850000
            dl1_cache_misses                                  = 55
            biu_reqs                                          = 19

          top.cpu.core0.lsu.tlb
              tlb_hits                                        = 5850001
```

I will further look into the stats to make sure that they make sense.
But the added stats can now be used for the purpose of the demo.

The next step is adding rename function to let instructions fly via
Scoreboard 🙂

Couple of issues fixed and cleaned up:

- Logger cleanup
- Fetch credit bug fix (fixes #23)
- Logger pure virtual call in Retire -- deleted inst pointer before done
with it!
- Added `isDone` to stop fetching when the trace is finished

In this PR, Execute is now it's own node and dynamically creates
execution pipes based on extensions listed in the
config files:
```
top.cpu.core0.extension.core_extensions:
  execution_topology:
  [["alu", "6"],
   ["fpu", "2"],
   ["br",  "2"]]
```

Work done;
- Added new class `Dispatcher` that is dynamically created by `Dispatch`
for each execution pipe defined in the `execution_topology` extension
- Added new class `Execute` that is a node created for the sole purpose
of dynamically creating the `ExecutePipe` objects based on the
`execution_topology` extension
- Cleaned up old code
- Reduced the `CPUTopology` structure to just static components like
Fetch, Decode, Rename, Dispatch, Execute, LS/etc, and Retire
- Updated the arches to include a small, medium, and big core. Dhrystone
trace improves with each architecture, but is limited by LS.

Signed-off-by: Knute Lingaard <knute.lingaard@sifive.com>
Co-authored-by: Zhen Wei <zhen.wei@sifive.com>

To facilitate dynamic execution topologies (what that means: being able
to add/subtract ALUs, branch units, FP units, etc), a Dispatch tester is
in order to ensure it's working properly.

This PR adds:

1. A Dispatch tester class
2. Generic Source/Sink classes for instruction groups or single
instruction
3. A more structured build environment
4. First cut of topology yaml files + core extensions